Thread: primary and hot standby database don' work now

hello everyone:
              my pg version is 9.1.2 running on red hat 5.4 x64 servers.one
day,i found all database indexes can't work,reindex database,it can work,but
some table has primary key,found exists the same two records in table.
             primary database error:
                         semop(id=13762606) failed: Identifier
removed",,,,,,"
                          ...................................
                        ERROR,XX000,"right sibling's left-link doesn't
match: block 6260 links to 6977 instead of expected 6398 in index
""tt_wait_artifident_ck_20120716_pk""",,,,,,"
                         ...............
                 WARNING,01000,"relation ""tt_telhis_src"" page 2390 is
uninitialized ---
                       ................
              hot standby:
                     2012-07-19 21:15:10.827
CST,,,19617,,5008085e.4ca1,1,,2012-07-19 21:15:10
CST,,0,LOG,00000,"streaming replication successfully connected to
primary",,,,,,,,,""

2012-07-19 21:18:36.133 CST,,,476,,4fde7dd3.1dc,22,,2012-06-18 09:01:07
CST,1/0,0,WARNING,01000,"specified item offset is too large",,,,,"xlog redo
insert: rel 17673/16386/16780; tid 409/151",,,,""

2012-07-19 21:18:36.133 CST,,,476,,4fde7dd3.1dc,23,,2012-06-18 09:01:07
CST,1/0,0,PANIC,XX000,"btree_insert_redo: failed to add item",,,,,"xlog redo
insert: rel 17673/16386/16780; tid 409/151",,,,""

2012-07-19 21:18:36.204 CST,,,474,,4fde7dd3.1da,4,,2012-06-18 09:01:07
CST,,0,LOG,00000,"startup process (PID 476) was terminated by signal 6:
Aborted",,,,,,,,,""

2012-07-19 21:18:36.204 CST,,,474,,4fde7dd3.1da,5,,2012-06-18 09:01:07
CST,,0,LOG,00000,"terminating any other active server processes",,,,,,,,,""

















--
View this message in context:
http://postgresql.1045698.n5.nabble.com/primary-and-hot-standby-database-don-work-now-tp5717562.html
Sent from the PostgreSQL - bugs mailing list archive at Nabble.com.

On 07/22/2012 02:00 PM, leo xu wrote:
> hello everyone:
>                my pg version is 9.1.2 running on red hat 5.4 x64 servers.one
> day,i found all database indexes can't work,reindex database,it can work,but
> some table has primary key,found exists the same two records in table.

When did this start happening? At random? After a server crash or power
loss? After an administrative action?

Are you running with fsync=off or using storage that has write caching
enabled?

--
Craig Ringer

On 22 July 2012 07:00, leo xu <leoxu8703@gmail.com> wrote:
> hello everyone:
>               my pg version is 9.1.2 running on red hat 5.4 x64 servers.one
> day,i found all database indexes can't work,reindex database,it can work,but
> some table has primary key,found exists the same two records in table.
>              primary database error:

> 2012-07-19 21:18:36.133 CST,,,476,,4fde7dd3.1dc,22,,2012-06-18 09:01:07
> CST,1/0,0,WARNING,01000,"specified item offset is too large",,,,,"xlog redo
> insert: rel 17673/16386/16780; tid 409/151",,,,""

There's something badly wrong with your WAL records.

I'd suggest you move to a later version of 9.1 and retry.

--
 Simon Riggs                   http://www.2ndQuadrant.com/
 PostgreSQL Development, 24x7 Support, Training & Services